The 2003 Institute of Medicine report, Unequal Treatment, proposed that racial attitudes contribute to Black- White health care disparities, which in turn result in health disparities. The goal of this research is to test a theoretical model addressing: (1) how Black patients' and nonBlack physicians' racial attitudes influence health care and (2) why some Black patients are more vulnerable to health care disparities than others. The model focuses on one aspect of health care: medical interactions. About 75% of Black patients experience racially discordant medical interactions with nonBlack physicians, and these interactions tend to be less positive than racially concordant ones. Prior research in social psychology provides strong evidence that racial attitudes play an important role in determining the quality of racially discordant interactions. Thus, the theoretical model proposed in this research posits that Black patients' and nonBlack physicians' racial attitudes (i.e., perceived discrimination and racial bias, respectively) negatively influence their affect and communication style during medical interactions; these then lead to decreased patients' adherence, which results in poorer health. The model also concerns variations in the quality of medical interactions among Black patients. The PI and others have shown that strong Afrocentric facial features (AFF; i.e., thick lips, wide nose) are associated with more perceived discrimination among Blacks and with more bias and stereotype activation among nonBlacks. Thus, the model posits that Black patients with stronger AFF display and elicit (in nonBlack physicians) more negative affect and communication style during medical interactions than Blacks with weaker AFF, and thus these patients have poorer adherence and health. We propose a secondary analysis of existing data that include video-recorded primary care interactions between Black patients and nonBlack physicians, patients' and physicians' reactions to the interactions, and longitudinal data on patients' adherence and health. Our aims are: Aim #1) To apply and test the validity of measures (i.e., observer judgments of thin slices and linguistic analysis of the interactions) of patients' and physicians' affect and communication style; and Aim #2) To empirically test a theoretical model that explains: (1) how patients' and physicians' racial attitudes influence the quality of racially discordant medical interactions, which in turn influence patients' subsequent health-related behaviors and health; and (2) how AFF influences racial attitudes, the quality of racially discordant medical interactions, and patients' subsequent adherence and health. Our study is significant, because it will inform the development of effective and efficient interventions addressing health disparities by focusing on relatively easily modifiable factors (affect, communication style) and specific at-risk patients (those with stronger AFF). This research is innovative because the model simultaneously considers the independent effects of Black patients' and nonBlack physicians' racial attitudes, addresses within-group variations in health care/health disparities, and is tested through multiple measures to assess affect and communication style.